In recent years, various types of three-dimensional (3D) image display apparatuses for viewing three-dimensional (3D) images, for instance, a three-dimensional (3D) TV, have been provided. In such 3D image display apparatuses, for example, a user is enabled to perceive 3D images (stereoscopic images) by using left-eye images and right-eye images based on binocular disparity.
Methods of displaying 3D images are classified into a glasses method using 3D glasses and a naked-eye stereopsis method (glasses-less 3D method). The methods of using 3D glasses include a shutter method (active method) and a polarization method (passive method). As the naked-eye stereopsis method (glasses-less 3D method), there are known, for instance, a lenticular method, a barrier method and an integral imaging method.
In these 3D image display apparatuses, an observer (user) can perceive normal stereoscopic images, that is, real stereoscopic images (“3D viewing”) when viewing the display screen from a correct position. However, if the viewing position deviates from the correct position, stereopsis is no longer effected, and a phenomenon, such as crosstalk (3D crosstalk) or pseudoscopy, will occur. If 3D crosstalk occurs, a decrease in stereoscopic effect of the images, a decrease in depth effect of the images or appearance of double images (ghost images) will occur. Pseudoscopy is a phenomenon in which a left-eye image is incident on the right eye and a right-eye image is incident on the left eye. The crosstalk and pseudoscopy may possibly become factors which cause eyestrain of the user.
However, even when the user views an ordinary 3D moving picture such as a 3D movie or 3D TV broadcast program, the user cannot always confirm whether the user is actually enjoying 3D viewing. The reason for this is that picture patterns vary one after another in an ordinary 3D moving picture, and the picture patterns are complex. Thus, the user is unable to discriminate whether stereoscopic images, which the user currently perceives, are real stereoscopic images, and it is difficult for the user to confirm whether the present viewing position is within a correct viewing area (3D viewing area).